Role of tyrosine kinase activity in α-adrenergic inhibition of the β-adrenergically regulated L-type Ca²⁺ current in guinea-pig ventricular myocytes

Abstract

1. The purpose of this study was to investigate the hypothesis that tyrosine kinase activity contributes to α₁-adrenergic inhibition of β-adrenergic responses in cardiac myocytes. We addressed this question by studying the pharmacological regulation of the L-type Ca²⁺ current in acutely isolated adult guinea-pig ventricular myocytes using the whole-cell patch-clamp technique.

2. The selective α₁-adrenergic receptor agonist methoxamine had no effect on the basal L-type Ca²⁺ current. Methoxamine also had no effect on cAMP-dependent stimulation of the Ca²⁺ current mediated by H₂ histamine receptor activation. However, methoxamine did inhibit cAMP-dependent stimulation of the Ca²⁺ current mediated by β-adrenergic receptor activation. The ability of methoxamine to inhibit β-adrenergic regulation of the Ca²⁺ current was significantly antagonized by the tyrosine kinase inhibitors genistein and lavendustin A.

3. The inhibitory effect of methoxamine was also mimicked by the phosphotyrosine phosphatase inhibitor pervanadate (PVN). PVN had no effect on basal Ca²⁺ current or Ca²⁺ current stimulated by histamine, but it did inhibit Ca²⁺ current stimulated by β-adrenergic receptor activation. Furthermore, the ability of PVN to inhibit β-adrenergic stimulation of the Ca²⁺ current was antagonized by lavendustin A.

4. These results are consistent with the conclusion that in guinea-pig ventricular myocytes α-adrenergic inhibition of β-adrenergic responses involves a tyrosine kinase-dependent signalling pathway. The fact that methoxamine and PVN antagonized cAMP-dependent responses mediated by β-adrenergic, but not H₂ histamine, receptor activation suggests that the inhibitory effect of α-adrenergic stimulation and tyrosine kinase activity is at the level of the β-adrenergic receptor.

The ability of α-adrenergic receptor (α-AR) stimulation to antagonize β-adrenergic responses is well documented in non-cardiac as well as cardiac preparations. In non-cardiac cells, such effects are associated with α₂-AR stimulation (Bylund, 1992). However, in cardiac myocytes, antagonism of β-adrenergic responses involves α₁-ARs (Buxton & Brunton, 1985, 1986; Boutjdir et al. 1992; Barrett et al. 1993; Iyadomi et al. 1995; Hartmann et al. 1995; Oleksa et al. 1996; Chen et al. 1996; Hool et al. 1997). This is consistent with the fact that cardiac myocytes primarily express α₁-, not α₂-, ARs (Fedida et al. 1993; Terzic et al. 1993). The mechanism responsible for the antagonistic effect of α₁-AR stimulation in cardiac muscle involves inhibition of β-adrenergically stimulated cAMP levels (Watanabe et al. 1977; Buxton & Brunton, 1985, 1986; Barrett et al. 1993; Lemire et al. 1998). The fact that the inhibitory effect is on the cAMP signalling pathway explains why α₁-AR stimulation affects a wide range of β-adrenergic responses in the heart. This includes β-adrenergic regulation of chronotropic effects (Molderings & Schumann, 1989), L-type Ca²⁺ channel activity (Boutjdir et al. 1992; Chen et al. 1996), CFTR Cl⁻ channel activity (Iyadomi et al. 1995; Oleksa et al. 1996; Hool et al. 1997) and contractility (Hartmann et al. 1995).

The mechanism by which α₁-AR stimulation affects cAMP-dependent responses in cardiac myocytes is not well understood. Biochemical studies have suggested that α₁-AR stimulation may either inhibit cAMP production (Barrett et al. 1993) or enhance cAMP degradation (Buxton & Brunton, 1985). Functional studies suggest that the former is more likely. Activation of the cAMP-regulated Cl⁻ current by the β-AR agonist isoprenaline (Iso) can be antagonized by α₁-AR stimulation. However, α₁-agonists are unable to inhibit the Cl⁻ current activated by direct stimulation of adenylate cyclase with forskolin, suggesting that α₁-AR stimulation exerts its effect upstream of cAMP production (Iyadomi et al. 1995; Oleksa et al. 1996). Consistent with this conclusion, α₁-agonists are also unable to inhibit persistent activation of the Cl⁻ current in the presence of the non-hydrolysable GTP analogue GTPγS (Iyadomi et al. 1995).

The signalling pathway linking the α₁-AR to its site of action is also unclear. In cardiac muscle, α₁-AR stimulation is often associated with the activation of PKC (Fedida et al. 1993; Terzic et al. 1993). However, we previously demonstrated that α₁-adrenergic inhibition of the β-adrenergically regulated cardiac Cl⁻ current does not involve PKC (Oleksa et al. 1996). Another signalling pathway associated with α₁-AR stimulation involves regulation of tyrosine kinase activity (Zhong & Minneman, 1999). Furthermore, we previously reported evidence that basal tyrosine kinase activity inhibits β-adrenergic responses in cardiac myocytes (Hool et al. 1998). Therefore, the purpose of this study was to test the hypothesis that α₁-AR stimulation inhibits β-adrenergic regulation of cardiac responses via a tyrosine kinase-dependent mechanism. To address this question, we evaluated the ability of tyrosine kinase inhibition to block, and phosphotyrosine phosphatase (PTP) inhibition to mimic, the effects of α₁-adrenergic stimulation on the L-type Ca²⁺ current in isolated guinea-pig ventricular myocytes. Our results are consistent with the idea that α₁-AR stimulation inhibits β-adrenergic responses at least partially through a tyrosine kinase-dependent mechanism that acts at the level of the β-AR. Some parts of this work have been published in abstract form (Nulton-Persson et al. 2000).

METHODS

Cell isolation

Single ventricular myocytes were isolated from adult Hartley guinea-pigs using the modification of a method described previously (Hool et al. 1998). The methods used in this study were approved by the Institutional Animal Care and Use Committee at Case Western Reserve University. Animals were anaesthetized by intraperitoneal injection of pentobarbital (150 mg kg⁻¹). Hearts were then quickly excised and the coronary arteries perfused via the aorta with a solution containing (mm): 140 NaCl, 5.4 KCl, 2.5 MgCl₂, 1.5 CaCl₂, 11 glucose and 5.5 Hepes (pH 7.4). The heart was perfused with this solution for 5 min, nominally Ca²⁺-free solution for 5 min, and then nominally Ca²⁺ free solution containing ∼0.5 mg ml⁻¹ collagenase (class B, Boehringer Mannheim) for ∼30 min. The ventricles were then removed and minced in a modified Kraft-Bruhe solution (Isenberg & Klockner, 1982) containing (mm): 110 potassium glutamate, 10 KH₂PO₄, 25 KCl, 2 MgSO₄, 20 taurine, 5 creatine, 0.5 EGTA, 20 glucose and 5 Hepes (pH 7.4). Single cells were obtained by filtering through nylon mesh, resuspended in Ca²⁺-containing solution, and used on the day of isolation only.

Acquisition and analysis of electrophysiological data

The β-adrenergically regulated L-type Ca²⁺ current was studied using the conventional whole cell configuration of the patch clamp technique as described previously (Hool & Harvey, 1997; Hool et al. 1998). Patch pipettes (1-2 MΩ) were filled with an intracellular solution containing (mm): 130 CsCl, 20 tetraethylammonium chloride (TEA-Cl), 5 MgATP, 5 EGTA, 0.1 Tris-GTP and 5 Hepes (pH 7.2). Cells were bathed in a K⁺-free control extracellular solution containing (mm): 140 NaCl, 5.4 CsCl, 2.5 CaCl₂, 0.5 MgCl₂, 11 glucose and 5.5 Hepes (pH 7.4). Isolated myocytes were placed in a 0.5 ml chamber on the stage of an inverted microscope, where they were superfused with either control or drug-containing extracellular solution at a rate of 1-2 ml min⁻¹. Some experiments were also conducted by superfusing the voltage-clamped myocyte using a fast flow system that allowed solutions to be changed in less than 1 s. All experiments were performed at room temperature.

Macroscopic currents were recorded using an Axopatch 200 voltage-clamp amplifier (Axon Instruments, Foster City, CA, USA) and an IBM-compatible computer with a Digidata 1200 interface and pCLAMP software (Axon Instruments). Analog input was low pass Bessel filtered at 5 kHz and sampled at 10 kHz. The membrane potential was held at -80 mV. Applying a 50 ms prepusle to -30 mV immediately before each test pulse inactivated Na⁺ channels. Using K⁺-free solutions containing Cs⁺ and TEA eliminated K⁺ currents. Setting the Cl⁻ equilibrium potential equal to the test potential eliminated the cAMP-dependent Cl⁻ current from the Ca²⁺ current measurements. However, when present, activation of the Cl⁻ current explains the changes in steady state current observed during the prepulse to -30 mV in some experiments (see Fig. 1). The size of the Ca²⁺ current was defined by measuring the absolute magnitude of the peak inward Ca²⁺ current recorded during 100 ms test pulses to 0 mV. The mean amplitude of the basal Ca²⁺ current recorded from all cells used in this study was 843 ± 25.0 pA (n =207).

Figure 1. The α₁-AR agonist methoxamine selectively inhibits the L-type Ca²⁺ current recorded in the presence of the β-AR agonist isoprenaline (Iso).

A, representative time course of changes in magnitude of peak inward Ca²⁺ current following exposure to 30 nm Iso and Iso plus 30 μm methoxamine. Currents were elicited by depolarizing test pulses to 0 mV applied once every 5 s. B, example of current traces recorded at time points indicated in the experiment illustrated in A. C, cumulative results of experiments in which cells were exposed to 30 μm methoxamine alone (n = 8; mean basal current amplitude, 902 ± 144 pA) or 30 μm methoxamine following exposure to 30 nm isoprenaline (n = 24; mean basal current amplitude, 816 ± 85.3 pA). Responses were normalized to the magnitude of the baseline current recorded before exposure to any drug(s). For each experiment, the magnitude of the current recorded in the presence of methoxamine was compared to the magnitude of the current recorded in the absence of methoxamine in the same cell. Statistical significance was evaluated by a paired t test.

The magnitude of the inhibitory effects of methoxamine and PVN were determined by calculating the percentage inhibition of the Iso response observed in each cell and then averaging the results obtained for all cells exposed to the same set of conditions. Results are reported as the mean ± the standard error of the mean. Statistical significance of the response was then determined by using Student's paired t test. When comparing responses between two different groups of cells, an unpaired t test was used. When comparing responses between more than two groups, a one-way analysis of variance and a t test with Bonferroni correction were used. P values <0.05 were considered significant. Where indicated, the concentration dependence of drug effects (E) were fitted to a logistic equation: E = E_max/(1 + ([drug]/EC₅₀)ⁿ) using a non-linear least squares curve-fitting routine (SigmaPlot, SPSS Inc., Chicago, IL, USA). In this relationship, E_max is the magnitude of the response to a maximally effective concentration of drug, EC₅₀ is the concentration of drug producing a half-maximal response, and n is the apparent Hill coefficient.

Immunoprecipitation of tyrosine phosphorylated proteins

Cell lysates were prepared by resuspending isolated myocytes in ice-cold lysis buffer containing: 100 mm Tris-HCl (pH 7.4), 100 mm NaCl, 10 mm EGTA, 10 mm NaF, 1 % Triton X-100, 2 mm sodium orthovanadate, 0.1 mm ammonium molybdate, 200 nm phenylarsine oxide, 2 μg ml⁻¹ leupeptin, and 2 μg ml⁻¹ aprotinin. Nuclei and other debris were removed by centrifugation first at 300 g for 10 min and then 100 000 g for 30 min. Lysates were precleared with normal mouse IgG, following which equal amounts of protein obtained from control myocytes or myocytes pretreated with 100 μm PVN for 5 min were subjected to immunoprecipitation overnight with antiphosphotyrosine antibody preconjugated to agarose beads (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA). The total immunoprecipitate obtained from each sample was separated on a 10 % polyacrylamide gel. After electrophoresis, the proteins were electrophoretically transferred to PVDF membrane and blocked by incubating for 1-2 h at room temperature in immunoblot buffer containing 10 mm Tris-HCl (pH 7.4), 0.9 % NaCl, 0.05 % Tween-20, 1 mm EDTA and 1% albumin. After blocking, membranes were then incubated for 1 h at room temperature in immunoblot buffer containing antiphosphotyrosine antibody conjugated to horseradish peroxidase (HRP). Immune reactive bands were then identified using an enhanced chemiluminescence (ECL) detection kit (Pierce, Rockford, IL, USA).

Drugs and chemicals

Iso, methoxamine and histamine were prepared as aqueous stock solutions and later diluted in extracellular solution to achieve the desired final concentration. Ascorbic acid (50 μm) was added to extracellular solutions to maintain the stability of Iso. Genistein (Alexis Pharmaceuticals), lavendustin A and lavendustin B (Calbiochem) were prepared as stock solutions in dimethylsulfoxide (DMSO). The final concentration of DMSO in extracellular solutions was 0.05-0.1%. DMSO alone in this concentration range had no effect on Iso or methoxamine responses. PVN was prepared as previously described (Hool et al. 1998), by combining 10 mm H₂O₂ and 10 mm Na₃VO₄ in an aqueous solution containing 50 mm Hepes (pH 7.4). This mixture was allowed to stand at room temperature for 15 min, after which time excess H₂O₂ was eliminated by adding catalase. The resulting stock solution contained a mixture of vanadate and peroxovanadate complexes (Posner et al. 1994). The final concentration of PVN used in our experiments is based on the concentration of Na₃VO₄ used in preparing the stock solution. It was verified that catalase and H₂O₂ alone had no effect on agonist responses (see Fig. 7). It was also verified that PVN did not result in the oxidative degradation of Iso by directly measuring its activity via electrochemical detection following separation on a reverse phase HPLC column (C. Sims and R. D. Harvey, unpublished observations). All solutions containing Iso and PVN were stored in light-resistant containers. All drugs were obtained from Sigma/RBI (St Louis, MO, USA), except where noted.

Figure 7. The phosphotyrosine phosphatase inhibitor pervanadate (PVN) selectively inhibits the L-type Ca²⁺ current recorded in the presence of the β-AR agonist isoprenaline (Iso).

A, representative time course of changes in magnitude of peak inward Ca²⁺ current following exposure to 30 nm Iso, Iso plus the vehicle for PVN (H₂O₂/catalase), and Iso plus 100 μm PVN. Currents were elicited by depolarizing test pulses to 0 mV applied once every 5 s. B, example of current traces recorded at time points indicated in the experiment illustrated in A. C, cumulative results of experiments in which cells were exposed to 100 μm PVN alone (n = 8; mean basal current amplitude, 830 ± 162 pA) or 100 μm PVN following exposure to 30 nm isoprenaline (n = 13; mean basal current amplitude, 1021 ± 88.84 pA). Responses were normalized to the magnitude of the baseline current recorded before exposure to any drug(s). For each experiment, the magnitude of the current recorded in the presence of PVN was compared to the magnitude of the current recorded in the absence of PVN in the same cell. Statistical significance was evaluated by a paired t test.

RESULTS

α₁-Adrenergic inhibition of β-adrenergic responses

The ability of α₁-AR stimulation to inhibit β-adrenergic responses is illustrated in Fig. 1. Exposure to the selective α₁-AR agonist methoxamine alone had no appreciable effect on basal Ca²⁺ channel activity. The magnitude of the current recorded following exposure to 30 μm methoxamine alone for 5 min decreased by 8 ± 3.7 % (n = 8). The small decrease can be explained by current rundown. This conclusion is supported by the observation that the magnitude of the current continued to gradually decline even after methoxamine was washed out (data not shown). However, if the L-type Ca²⁺ current was first stimulated by exposure to the β-adrenergic agonist Iso, methoxamine exerted a significant inhibitory effect. Exposure to 30 nm Iso alone increased the Ca²⁺ current by 256 ± 26.8 % (n = 24) over baseline. Subsequent exposure to 30 μm methoxamine inhibited the response to Iso by 47 ± 7.0 %.

The ability of methoxamine to inhibit β-adrenergic regulation of the L-type Ca²⁺ current via activation of α₁-ARs is consistent with previous work demonstrating that the ability of methoxamine to inhibit β-adrenergic regulation of ion channel activity as well as contraction can be antagonized by the selective α₁-AR antagonist prazosin (Molderings & Schumann, 1989; Hartmann et al. 1995; Oleksa et al. 1996). We also previously demonstrated that the ability of α₁-adrenergic stimulation to inhibit the cAMP-regulated Cl⁻ current activated by a given concentration of Iso can be overcome by increasing the level of β-AR stimulation (Hool et al. 1997). This suggests that α₁-AR stimulation inhibits β-adrenergic responses in a competitive manner. To ascertain whether the same mechanism is involved in α₁-adrenergic inhibition of the β-adrenergically regulated Ca²⁺ current, we studied the effect that methoxamine has on the concentration dependence of Iso-induced stimulation of the Ca²⁺ current (Fig. 2). In the absence of methoxamine, Iso stimulated the Ca²⁺ current with an EC₅₀ of 2.5 ± 0.26 nm. However, the EC₅₀ for Iso increased to 14 ± 5.6 nm in the presence of 30 μm methoxamine (P < 0.05). These results support the idea that α₁-AR stimulation inhibits β-adrenergic activation of the L-type Ca²⁺ current in a competitive manner.

Figure 2. The α₁-AR agonist methoxamine decreases the sensitivity of the L-type Ca²⁺ current to β-AR stimulation.

At each concentration of isoprenaline, the amplitude of the Ca²⁺ current was measured in the absence (▪) and presence (□) of 30 μm methoxamine. These values are presented as the percentage increase over the magnitude of the basal Ca²⁺ current measured in the same cell in the absence of any drug. The continuous lines represent the best fits of the respective data points to a logistic equation (see Methods). Control: E_max= 246 ± 8.32 %; EC₅₀= 2.5 ± 0.26 nm; n = 2.4± 0.54. Methoxamine: E_max= 243 ± 28.1 %; EC₅₀= 14 ± 5.6 nm; n = 1.1± 0.40. The numbers in parentheses indicate sample population size.

Lack of α₁-adrenergic inhibition of histamine responses

The inhibitory effects of α₁-agonists on β-adrenergic responses can be explained by the ability of α₁-AR stimulation to antagonize β-AR-induced production of cAMP in cardiac myocytes (Watanabe et al. 1977; Buxton & Brunton, 1985, 1986; Barrett et al. 1993; Lemire et al. 1998). To take this one step further, we previously demonstrated that in guinea-pig ventricular myocytes, methoxamine inhibits the Cl⁻ current activated by Iso, but it has little or no effect on the Cl⁻ current activated by histamine (Oleksa et al. 1996). Histamine regulates cardiac responses through the same cAMP-dependent signalling pathway that is linked to β-ARs, except it does so via the activation of H₂ histamine receptors (Hescheler et al. 1987). This suggests that α₁-adrenergic stimulation inhibits β-adrenergic responses by acting at the level of the β-AR to inhibit cAMP production. To verify that α₁-adrenergic inhibition of the β-adrenergically activated Ca²⁺ current was due to the same type of mechanism, we examined the effect of methoxamine on the L-type Ca²⁺ current activated by histamine.

Exposing cells to 300 nm histamine enhanced the L-type Ca²⁺ current by 214 ± 20.7 % (n = 12) over baseline, and subsequent exposure to methoxamine had no obvious inhibitory effect (Fig. 3). During exposure to 30 μm methoxamine in the continued presence of histamine the magnitude of the total current decreased by 5.8 ± 2.2 %. This small decrease can be explained by current rundown. Consistent with this conclusion, subsequent washout of methoxamine did not result in an increase in the current, as would be expected if exposure to methoxamine had caused a true inhibitory effect. The fact that the relative change in total current magnitude observed during exposure to 30 μm methoxamine was independent of whether the current was first stimulated by histamine supports the idea that α₁-AR stimulation is unable to inhibit histamine responses.

Figure 3. The α₁-AR agonist methoxamine does not inhibit the L-type Ca²⁺ current recorded in the presence of histamine.

A, representative time course of changes in magnitude of peak inward Ca²⁺ current following exposure to 300 nm histamine and histamine plus 30 μm methoxamine. Currents were elicited by depolarizing test pulses to 0 mV applied once every 5 s. B, example of current traces recorded at time points indicated in the experiment illustrated in A. C, cumulative results of experiments in which cells were exposed to 30 μm methoxamine following exposure to 300 nm histamine (n = 12; mean basal current amplitude, 854 ± 138 pA). Responses were normalized to the magnitude of the baseline current recorded before exposure to histamine. For each experiment, the magnitude of the current recorded in the presence of methoxamine was compared to the magnitude of the current recorded in the absence of methoxamine in the same cell. Statistical significance was evaluated by a paired t test.

Because the ability of α₁-AR stimulation to inhibit β-adrenergic responses is competitive, the inability of methoxamine to inhibit the histamine response might be explained if the level of stimulation produced by 300 nm histamine were greater than that produced by 30 nm Iso. To test this possibility, we studied the effect of methoxamine on the concentration dependence of histamine-induced stimulation of the Ca²⁺ current (Fig. 4). In the absence of methoxamine, histamine stimulated the Ca²⁺ current with an EC₅₀ of 40 ± 2.6 nm. This is not significantly different (P > 0.5) from the EC₅₀ of 38 ± 5.7 nm observed in the presence of methoxamine. This result supports the conclusion that histamine regulation of the Ca²⁺ current is not affected by stimulation of α₁-ARs with methoxamine. Consistent with this idea, increasing the concentration of methoxamine 10-fold did not have a significantly greater effect on the response to histamine (data not shown). The magnitude of the total current measured in the presence of 300 nm histamine decreased by 7.4 ± 2.1 % (n = 7) following addition of 300 μm methoxamine. Again, this small decrease can be attributed largely to current rundown. Although in some experiments there was evidence for a small inhibitory effect that was reversible upon washout, it is unlikely that this was due to antagonism of the histamine response, since exposure to 300 μm methoxamine in the absence of histamine caused a similar small but reversible decrease of the basal Ca²⁺ current in some cells. The magnitude of the current measured in the presence of 300 μm methoxamine alone decreased to 77 ± 3.5 % (n = 7) of that observed before exposure to the drug, and following washout the current returned to 87 ± 2.8 % of its original size.

Figure 4. The α₁-AR agonist methoxamine has no effect on the sensitivity of the L-type Ca²⁺ current to histamine receptor activation.

At each concentration of histamine, the amplitude of the Ca²⁺ current was measured in the absence (▪) and presence (□) of 30 μm methoxamine. These values are presented as the percentage increase over the magnitude of the basal Ca²⁺ current measured in the same cell in the absence of any drug. The continuous lines represent the best fits of the respective data points to a logistic equation (see Methods). Control: E_max= 225 ± 8.39 %; EC₅₀= 40 ± 2.6 nm; n = 3.8± 0.71. Methoxamine: E_max= 223 ± 19.3 %; EC₅₀= 38 ± 5.7 nm; n = 4.0± 1.9. The numbers in parenthesis indicate sample population size.

Tyrosine kinase inhibitors antagonize the α₁-adrenergic response

The fact that β-adrenergic responses are selectively inhibited by methoxamine suggests that α₁-adrenergic stimulation may act either directly or indirectly at the level of the β-AR producing a desensitization-like effect. We previously demonstrated that the drug genistein increases the sensitivity of cAMP-regulated Cl⁻, Ca²⁺, and K⁺ channels to β-adrenergic stimulation (Hool et al. 1998). Genistein is a selective tyrosine kinase inhibitor (Akiyama & Ogawara, 1991). This suggests that tyrosine phosphorylation can exert an inhibitory effect on the β-adrenergic signalling pathway in these cells. Therefore, we evaluated the possibility that α₁-adrenergic inhibition of β-adrenergic responses might involve tyrosine phosphorylation. To test this hypothesis, we examined the effect of methoxamine on responses to Iso in cells pretreated with genistein (Fig. 5). Exposure to 50 μm genistein alone inhibited the basal Ca²⁺ current by 42 ± 5.6 % (n = 7). This has been described previously and may be explained by a direct blocking effect of genistein that is independent of inhibition of tyrosine kinase activity (Hool et al. 1998). Nonetheless, exposure to 30 nm Iso in the continued presence of genistein still increased the magnitude of the current 302 ± 33.1 % (n = 7) over baseline, which is similar to the magnitude of the response to Iso observed in the absence of genistein. Under these conditions, 30 μm methoxamine inhibited the response to 30 nm Iso by 21 ± 4.5 %. This is significantly smaller (P < 0.05) than the 47 % decrease in the magnitude of the response to 30 nm Iso produced by the same concentration of methoxamine in the absence of genistein.

Figure 5. The tyrosine kinase inhibitor genistein attenuates the ability of the α₁-AR agonist methoxamine to inhibit the L-type Ca²⁺ current recorded in the presence of isoprenaline (Iso).

A, representative time course of changes in magnitude of peak inward Ca²⁺ current following exposure to 30 nm Iso and Iso plus 30 μm methoxamine in the presence of 50 μm genistein. Currents were elicited by depolarizing test pulses to 0 mV applied once every 5 s. B, example of current traces recorded at time points indicated in the experiment illustrated in A. C, magnitude of methoxamine-induced inhibition of Iso response in the absence (n = 24; mean basal current amplitude, 816 ± 85.3 pA) and presence (n = 7; mean basal current amplitude, 401 ± 45.6 pA) of 50 μm genistein. Statistical significance was evaluated by an unpaired t test.

It was not possible to use daidzein, the less active structural analogue of genistein, as a negative control, because under the conditions used in these experiments (i.e. room temperature) daidzein was not sufficiently soluble in the extracellular solution. Therefore, to further test the hypothesis that α₁-adrenergic inhibition of β-adrenergic responses involves a tyrosine kinase-dependent mechanism we studied the effects of lavendustin A, another tyrosine kinase inhibitor (Onoda et al. 1989). Unlike genistein, lavendustin A had no obvious effect on the basal Ca²⁺ current. The magnitude of the current recorded following exposure to 5 μm lavendustin A alone for 5 min decreased by 9.3 ± 3.7 % (n = 8), which is consistent with current rundown, independent of any drug effect. This is supported by the fact that there was no increase in current magnitude following washout of the drug, as would be expected if lavendustin A had produced a genistein-like non-specific effect. Furthermore, lavendustin A did not appear to have any obvious effect on the response to 10 nm Iso. In the absence of lavendustin A, 10 nm Iso increased the magnitude of the Ca²⁺ current 250 ± 32.4 % over baseline. In the presence of 5 μm lavendustin A, 10 nm Iso increased the magnitude of the Ca²⁺ current 233 ± 15.6 % over baseline.

Next, we compared the effect of 30 μm methoxamine on the response to 10 nm Iso with and without pre-exposure to 5 μm lavendustin A (Fig. 6). In the absence of lavendustin A, 30 μm methoxamine inhibited the response to 10 nm Iso by 59 ± 6.5 % (n = 13). However, in the presence of lavendustin A, methoxamine inhibited the response to 10 nm Iso by only 31 ± 6.6 % (n = 13). This represents a significant reduction in the magnitude of the response to methoxamine (P < 0.01). In the presence of 5 μm lavendustin B, a structurally related compound that only weakly inhibits tyrosine kinase activity, methoxamine inhibited the response to 10 nm Iso by 45 ± 5.0 % (n = 11). This is not significantly different from the magnitude of the response to methoxamine observed in the absence of lavendustin (P > 0.1). This is consistent with the idea that the α-adrenergic agonist methoxamine inhibits β-adrenergic regulation of the L-type Ca²⁺ current via a mechanism that is sensitive to inhibition of tyrosine kinase activity. Although the average size of the methoxamine response observed in the presence of lavendustin B was greater than that observed in the presence of lavendustin A, the difference did not achieve statistical significance (P > 0.05). This might be explained if lavendustin B were able to at least partially affect the activity of the tyrosine kinase involved (Onoda et al. 1989).

Figure 6. The tyrosine kinase inhibitor lavendustin A attenuates the ability of the α₁-AR agonist methoxamine to inhibit the L-type Ca²⁺ current recorded in the presence of isoprenaline (Iso).

A, representative time course of changes in magnitude of peak inward Ca²⁺ current following exposure to 10 nm isoprenaline and isoprenaline plus 30 μm methoxamine in the presence of 5 μm lavendustin A. Currents were elicited by depolarizing test pulses to 0 mV applied once every 5 s. B, example of current traces recorded at time points indicated in the experiment illustrated in A. C, magnitude of methoxamine-induced inhibition of Iso response in the absence of any other drug (n = 9; mean basal current amplitude, 817 ± 114 pA), in the presence of 5 μm lavendustin A (n = 13; mean basal current amplitude, 951 ± 94.5 pA), and in the presence of the inactive analogue lavendustin B (n = 11; mean basal current amplitude, 853 ± 105 pA). Statistical significance was evaluated by one-way analysis of variance and a t test with Bonferroni correction.

PVN mimics the α₁-adrenergic response

If the ability of α₁-adrenergic stimulation to inhibit β-adrenergic responses is due to an increase in tyrosine phosphorylation and genistein and lavendustin A block this response by inhibiting the kinase responsible, then one might predict that it would be possible to mimic the effects of methoxamine by inhibiting PTP activity, as long as the kinase involved is basally active. To test this possibility we studied the effect of the PTP inhibitor PVN on Iso-dependent regulation of the L-type Ca²⁺ current (Fig. 7). PVN alone had no apparent effect on the basal current. The magnitude of the Ca²⁺ current recorded following exposure to 100 μm PVN decreased by 7 ± 1.2 % (n = 8). Again, this small decrease in current magnitude can be explained by rundown since the current continued this gradual decline even after PVN was washed out. Despite having no obvious effect on basal channel activity, PVN had a significant effect on the Ca²⁺ current in the presence of Iso. Following exposure to 30 nm Iso, 100 μm PVN inhibited the total current magnitude by 54 ± 2.0 % (n = 13). This represents an 87 ± 3.1 % inhibition of the response to Iso. The observation that the change in current magnitude during exposure to PVN was more significant (P < 0.001) in the presence of Iso than it was in the absence of Iso is consistent with the idea that inhibition of PTP activity is able to antagonize β-adrenergic regulation of the Ca²⁺ current.

The ability of PVN to inhibit the Ca²⁺ current in the presence but not the absence of Iso is also consistent with the idea that α₁-AR stimulation may inhibit β-adrenergic responses through a tyrosine kinase-dependent mechanism. In an attempt to draw a stronger correlation between the role of tyrosine phosphorylation and the inhibitory effects of α₁-AR stimulation we next looked at the effects of PVN on histamine-dependent regulation of the Ca²⁺ current (Fig. 8). If PVN inhibition of the response to Iso were a non-specific effect, one might predict that it would inhibit all cAMP-dependent responses. However, if PVN were inhibiting Iso responses through the same mechanism that is involved in α₁-adrenergic inhibition of β-adrenergic responses, then it should have little or no effect on the response to an equivalent concentration (approx. 10 times the EC₅₀) of histamine. Consistent with this latter idea, PVN did not have a significant effect on the Ca²⁺ current stimulated by histamine. Following stimulation of the L-type Ca²⁺ current with 300 nm histamine, exposure to 100 μm PVN was associated with a decrease in the total current magnitude of 12 ± 4.0 % (n =6). The slight decrease can again be attributed in large part to current rundown, although in one cell exposure to PVN was associated with a small decrease in current magnitude that reversed upon washout. Overall, however, the change in total current magnitude observed during exposure to 100 μm PVN was independent (P > 0.1) of whether the current was first stimulated by histamine. This supports the idea that inhibition of PTP activity is unable to significantly inhibit histamine responses.

Figure 8. The phosphotyrosine phosphatase inhibitor pervanadate (PVN) does not inhibit the L-type Ca²⁺ current recorded in the presence of histamine.

A, representative time course of changes in magnitude of peak inward Ca²⁺ current following exposure to 300 nm histamine and histamine plus 100 μm PVN. Currents were elicited by depolarizing test pulses to 0 mV applied once every 5 s. B, example of current traces recorded at time points indicated in the experiment illustrated in A. C, cumulative results of experiments in which cells were exposed to 100 μm PVN in the presence of 300 nm histamine (n = 6; mean basal current amplitude, 889 ± 140 pA). Responses were normalized to the magnitude of the baseline current recorded before exposure to any drug(s). Statistical significance was evaluated by a paired t test.

The ability of tyrosine phosphatase inhibitors to selectively antagonize β-adrenergic regulation of cardiac ion channel function confirms our previous results (Sims et al. 2000). To take this further, we verified that PVN is actually causing an increase in tyrosine phosphorylation of cellular proteins by using an anti-phosphotyrosine antibody to immunoprecipitate tyrosine phosphorylated proteins from membrane preparations obtained from control myocytes and myocytes treated with 100 μm PVN for 5 min. Western blot of these precipitated proteins with the same antibody demonstrates that exposure to PVN, for the same period of time that produces functional responses, clearly increased the level of protein tyrosine phosphorylation in cardiac myocytes (Fig. 9). This observation is consistent with the idea that the effect of PVN on β-adrenergic regulation of Ca²⁺ channel function is associated with a tyrosine phosphorylation-dependent mechanism. This conclusion is further supported by the finding that the ability of PVN to inhibit β-adrenergic regulation of the L-type Ca²⁺ current was antagonized by lavendustin A (Fig. 10). In the presence of lavendustin A, 100 μm PVN inhibited the response to 30 nm Iso by only 38 ± 6.7 % (n = 10). This is significantly less than the 87 % inhibition (see Fig. 7) observed in the absence of lavendustin (P < 0.001). However, in the presence of lavendustin B, 100 μm PVN inhibited the response to 30 nm Iso by 71 ± 8.3 % (n = 9). This is significantly different from the magnitude of the inhibitory response that was observed in the presence of lavendustin A (P < 0.002), but it is not significantly different from the magnitude of the inhibitory effect that was observed in the absence of either lavendustin derivative (P > 0.1).

Figure 9. Pervanadate-induced increase in tyrosine phosphorylation of proteins in guinea-pig ventricular myocytes.

Anti-phosphotyrosine antibody was used to immunoprecipitate (IP) tyrosine-phosphorylated proteins from cell lysates prepared from untreated (control) myocytes and myocytes exposed to 100 μm pervanadate for 5 min (PVN treated). Similar results were obtained in a total of 3 experiments. Detection of tyrosine-phosphorylated proteins in the control lane was minimized due to the brief ECL exposure time, which was used to minimize saturation of signals in the PVN-treated lane.

Figure 10. The tyrosine kinase inhibitor lavendustin A attenuates the ability of the phosphotyrosine phosphatase inhibitor pervanadate (PVN) to inhibit the L-type Ca²⁺ current recorded in the presence of isoprenaline (Iso).

A, representative time course of changes in magnitude of peak inward Ca²⁺ current following exposure to 30 nm Iso and Iso plus 100 μm PVN in the presence of lavendustin A (5 μm external; 0.5 μm internal). Currents were elicited by depolarizing test pulses to 0 mV applied once every 5 s. B, example of current traces recorded at time points indicated in the experiment illustrated in A. C, magnitude of PVN-induced inhibition of Iso response in the absence of any other drug (n = 13; mean basal current amplitude, 1021 ± 88.84 pA), in the presence of lavendustin A (5 μm external; 0.5 μm internal) (n = 10; mean basal current amplitude, 819 ± 50.3 pA), and in the presence of the inactive analogue lavendustin B (5 μm external; 0.5 μm internal) (n = 9; mean basal current amplitude, 940 ± 98.5 pA). Statistical significance was evaluated by one-way analysis of variance and a t test with Bonferroni correction.

DISCUSSION

In the present study, we demonstrated that relatively low concentrations of the selective α₁-adrenergic agonist methoxamine had little or no effect on basal Ca²⁺ current. This is consistent with previous reports demonstrating that α₁-AR agonists have no appreciable effect on the basal L-type Ca²⁺ current in adult mammalian myocytes (Fedida et al. 1993; Terzic et al. 1993). However, we did find that methoxamine significantly inhibited the Ca²⁺ current that had been pre-stimulated by the β-adrenergic agonist Iso. This is in agreement with previous reports demonstrating that activation of α₁-ARs antagonizes various β-adrenergic responses, including β-adrenergic-induced chronotropic effects in guinea-pig atria (Molderings & Schumann, 1989), β-adrenergic stimulation of L-type Ca²⁺ channel activity in rat ventricular myocytes (Boutjdir et al. 1992; Chen et al. 1996), β-adrenergic activation of Cl⁻ channel activity in guinea-pig ventricular myocytes (Iyadomi et al. 1995; Oleksa et al. 1996; Hool et al. 1997), and β-adrenergic-induced changes in myocardial contractility and protein phosphorylation in rat hearts (Hartmann et al. 1995). Furthermore, cardiac-specific overexpression of α_1B-ARs in mice has been found to attenuate β-adrenergic responsiveness (Akhter et al. 1997; Lemire et al. 1998).

The ability of α₁-AR stimulation to inhibit functional responses to β-AR stimulation can be explained by the ability of α₁-AR activation to antagonize β-adrenergically stimulated cAMP levels. Buxton & Brunton (1985) originally suggested that the decrease in cAMP accumulation associated with α₁-AR stimulation was due to an increase in cAMP breakdown. This was based on the loss of the α₁-AR-dependent response in the presence of the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine (IBMX). However, Barrett et al. (1993) found that IBMX did not prevent changes in cAMP associated with α₁-AR stimulation, and concluded that α₁-AR stimulation inhibits cAMP synthesis. Functional studies would appear to support the latter idea. Boutjdir et al. (1992) found that α₁-AR stimulation had no effect on the L-type Ca²⁺ current activated directly by introduction of cAMP through the patch pipette, suggesting that α₁-AR stimulation did not exert an inhibitory effect directly at the level of the PKA-activated channel. Furthermore, Iyadomi et al. (1995) demonstrated that while α₁-adrenergic stimulation could inhibit cAMP-regulated Cl⁻ current stimulated by Iso in guinea-pig myocytes, it had no effect on the Cl⁻ current stimulated by direct activation of adenylate cyclase with forskolin, and it had no effect on the Cl⁻ current persistently activated by Iso in the presence of GTPγS. This suggested that the inhibitory effect of α₁-AR stimulation was acting upstream of adenylate cyclase activation. Our earlier work took this one step further, and demonstrated that α₁-AR stimulation did not inhibit the Cl⁻ current activated by histamine (Oleksa et al. 1996). Although histamine elicits cAMP responses identical to those produced by Iso, it does so through the activation of the H₂ histamine receptor. This suggested that the inhibitory effect of α₁-AR stimulation acts specifically and selectively at the level of the β-AR. The results of the present study confirm this conclusion by demonstrating that methoxamine inhibits cAMP-dependent stimulation of the L-type Ca²⁺ current by Iso, but not by histamine.

Although the results of previous studies suggested that α₁-AR stimulation acts directly at the level of the β-AR, there is much less information available concerning the signalling pathway that might lead to such an effect. We previously demonstrated that α₁-adrenergic inhibition of the β-adrenergically regulated Cl⁻ current involves a pertussis toxin-insensitive G protein (Hool et al. 1997). However, what effector this G protein is coupled to is unclear. In the heart, α₁-adrenergic responses are often associated with the activation of phospholipase C, production of diacylglycerol and stimulation of PKC, but this signalling pathway does not seem to be involved in α₁-adrenergic inhibition of β-adrenergic responses. We previously demonstrated that in guinea-pig ventricular myocytes, activation of endogenous PKC with a phorbol ester did not mimic the inhibitory effect of α₁-AR stimulation, and inhibition of PKC activity with bisindolylmaleimide did not block the α₁-adrenergic response (Oleksa et al. 1996).

Chen et al. (1996), on the other hand, reported that the PKC inhibitor calphostin C could partially attenuate α₁-AR-mediated inhibition of β-adrenergically stimulated unitary Ca²⁺ channel activity in rat ventricular myocytes. However, that response may reflect a direct inhibitory effect of PKC on Ca²⁺ channel activity. This idea is supported by the fact that in the same study, it was found that α₁-AR stimulation could inhibit cAMP-dependent Ca²⁺ channel activity produced independently of the β-AR. In addition, in a separate study, the same group demonstrated that activation of endogenous PKC with a phorbol ester inhibited basal Ca²⁺ current and β-adrenergically stimulated Ca²⁺ current similarly in rat cardiac myocytes (Zhang et al. 1997). The role of this type of direct PKC-dependent effect in mediating α₁-adrenergic responses is unclear. It seems unlikely to explain the inhibitory action that α₁-AR stimulation has on β-adrenergic responses, as described in present study, since relatively low concentrations of the α₁-AR agonist methoxamine had no effect on basal Ca²⁺ current and no effect on cAMP-dependent stimulation of the Ca²⁺ current by histamine.

Another signalling pathway associated with α₁-AR stimulation involves regulation of tyrosine kinase activity (Zhong & Minneman, 1999). It is well documented that stimulation of α₁-AR can also lead to activation of mitogenic responses in a variety of cell types including cardiomyocytes, and the tyrosine kinase inhibitor genistein has been shown to prevent α₁-AR-induced cardiac hypertrophy (Thorburn & Thorburn, 1994). Furthermore, in a previous study we demonstrated that genistein increases the sensitivity of the L-type Ca²⁺ current, as well as the cAMP regulated Cl⁻ current and the delayed rectifier K⁺ current, to β-AR stimulation (Hool et al. 1998). This suggests that basal tyrosine kinase activity exerts an inhibitory influence over the β-adrenergic responsiveness of cardiac myocytes. Our present results are consistent with the possibility that α₁-AR stimulation exerts its inhibitory effect on β-ARs at least partially via a tyrosine kinase-dependent mechanism. This includes the observation that the tyrosine kinase inhibitors genistein and lavendustin A significantly attenuated the inhibitory effect of methoxamine.

Furthermore, we found that PVN mimicked the effect of methoxamine, both in its ability to inhibit Iso-stimulated current and its inability to affect basal or histamine-stimulated current. Although such results might be explained if PVN were able to act as a selective β-AR antagonist, vanadate has been shown to have no effect on the affinity of β-ARs for Iso (Krawietz et al. 1982). Therefore, the effects of PVN on β-adrenergic regulation of the Ca²⁺ current are most likely to be associated with its ability to cause an increase in tyrosine phosphorylation of numerous cellular proteins (see Fig. 9). This conclusion is supported by the fact that the effects of PVN were significantly antagonized by lavendustin A, but not lavendustin B. Further similarity between the effects of α₁-AR agonists and PVN is supported by fact that PTP inhibitors in addition to PVN have been reported to antagonize β-adrenergic regulation of the Ca²⁺ current in guinea-pig ventricular myocytes in a competitive manner (Sims et al. 2000).

The question then is how might tyrosine phosphorylation affect the β-AR. One possibility is that there is direct tyrosine phosphorylation of the receptor protein. Karoor & Malbon (1998) have demonstrated that in non-cardiac preparations, β₂-ARs can be phosphorylated on tyrosine residues located in the second intracellular loop and the carboxy terminal tail. Tyrosine phosphorylation in either location results is an uncoupling of the receptor from the stimulatory G protein, G_s. However, we have previously demonstrated that Iso regulation of ion channel activity in guinea-pig ventricular myocytes is mediated solely through the activation of β₁-ARs (Hool & Harvey, 1997), and it has yet to be determined whether the β₁-AR is a substrate for tyrosine phosphorylation. Another possibility is that tyrosine phosphorylation affects β₁-AR function indirectly through the G protein-coupled receptor kinase GRK2, which regulates the function of β₁-ARs in the heart. Consistent with this idea, tyrosine phosphorylation has been found to upregulate the activity of GRK2 (Sarnago et al. 1999). Furthermore, cardiac-specific overexpression of α_1B-ARs upregulates GRK2 activity (Akhter et al. 1997). Therefore, there are at least two viable hypotheses as to how tyrosine phosphorylation might affect β₁-AR function in cardiac tissue. However, it is yet to be determined specifically whether α₁-AR stimulation causes tyrosine phosphorylation of either the β₁-AR or GRK2.